- Original Poster
- #1,241
And the Economist, previously among one of the more alarmist publications on the topic of Global Warming, just published an article revealing that global temperatures over the last 15 years have been stagnant - despite a change in CO2 levels.Here's a graph showing the rise of co2 in the past 50-60 years fro 305ppm to 395ppm
http://www.washingtonpost.com/blogs...their-highest-point-in-at-least-800000-years/
Phil: Another of my big gripes in this debate is my experience with forecasting models - both for near-term weather and long-time climate. During my PhD, I consumed 18 months of DEC-10 processing time and needed time on some of the big government Cray and Ciber machines to keep making progress. After many requests, I was eventually given an allotment of time on both. It turns out that by far the biggest users of these computers were the forecasters - yet their success rates were notoriously poor. Frankly, their models just couldn't take into account the very many factors that drive both weather and climate. Now, I understand we live in an age now with much more computer firepower, and models are more sophisticated; however, from those I know, there remains a lot of doubt and uncertainty in the field.
It's a very unfortunate and inconvenient fact that most models used to produce the famous hockey sticks over the last several years produce those graphs even with random input data. In other words, their starting assumptions lead to a predicted rise in temperatures no matter what the input data. Worse, the actual global temperatures witnessed in the last few years contradict all these models - as temperatures remain stagnant. That ought to tell us something.
Computer modeling is a difficult area of science. I remember once, when modeling fine particle magnets, predicting a peak in the specific heat. This generated a lot of excitement because it implied a new type of phase change. Thank goodness I ran the results past a brilliant colleague, because he spotted a problem in the modeling approach. I was wrong, the model was wrong, and double-checking against other data proved it. So, when actual data consistently fails to match predicted data, it's time to cross-check and re-evaluate.
Last edited:
Upvote
0
